JP2002528029A - Electronic commutation type motor - Google Patents

Abstract

In an electronically commutated motor (M), rotor position signals are generated by means of a galvanomagnetic rotor position sensor ( 40 ). A timer (CNT_HL) brings about an advanced commutation which occurs only once the motor has reached a specific rotation speed, and whose magnitude is a function of the rotation speed.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

【Technical field】

The present invention relates to an electronic commutation motor, and more particularly, to an electronic commutation motor that performs "early ignition". Early ignition is understood to mean that the commutation is shifted to an earlier point in time, usually as a function of the speed. Of course, it is not "ignited" in electric motors, but for its simplicity, this concept (borrowed from automotive technology) is actually preferred and is therefore called "ignition angle shift". Therefore, this concept is used below, even though this is not entirely scientifically correct.

[0002]

[Problems with conventional technology]

An electronic commutation type motor for performing early ignition is disclosed, for example, in DE-A197004.
79.2 (intern: D201i). Here, the accuracy of commutation is not sufficient for many cases, and the program must be executed according to a set time scheme. However, this is cumbersome, and the computing power of the processor is often used only poorly. The commutation process also varies considerably over time. This increases the noise of such motors.

[0003]

[Problems to be solved by the invention]

It is therefore an object of the present invention to provide a new electronic commutation motor and a method for driving such a motor.

[0004]

[Solution]

This object is achieved according to a first aspect of the invention by an electronic commutation motor according to claim 1. Such motors operate with relatively good efficiency, especially at high speeds. This is because the commutation can be shifted more quickly as the rotational speed increases. By using the interrupt routine, the commutation process can be precisely controlled in time, and thus the rotation of the motor becomes quiet.

Another solution to the problem is obtained by a method according to the invention. By measuring an additional second time following the passage of the first time, this 2
By adding one time and possibly a correction factor, it is very easy to obtain an amount of time that is substantially inversely proportional to the motor speed. This amount of time can be used in a subsequent commutation process to calculate a new value for the first time as an updated amount of time.

According to claim 17, this time amount is advantageously used for commutation processes in which the rotor rotation is after the measurement of the first and second times. This is because a particularly quiet rotation of the motor can be obtained in this case. For example, the amount of time is 0
This amount of time can be the basis for controlling commutation after one revolution, if measured in the rotation angle range from ゜ el to 180 ゜ el. This is because the commutation takes place in substantially the same angular range 0 ° el to 180 ° el.

[0007] It is particularly advantageous if the at least one non-time-critical method step is configured as a subroutine. This subroutine is called during the execution of the program when the processing time for it is available. This makes it possible to optimally use the resources of the processor, unlike programs with a fixed time scheme. This is because in this method, the subroutine is processed when the processor does not just do it.

[0008] Further details and advantageous refinements of the invention will become apparent from the following description and the embodiments illustrated in the drawings, as well as from the dependent claims. However, the examples should not be understood as limiting the invention.

Overview of Electronic Commutation Motor (ECM) FIG. 1 is a schematic diagram of a preferred embodiment of the electronic commutation motor (ECM) of the present invention. This motor is controlled by a microcontroller (μC) 11 or a microprocessor. ΜC11 (COP842 used in the examples)
CJ) terminals are shown by way of example in FIG.

A program executed by the μC 11 is structured by a function manager. This function manager will be described later with reference to FIGS.

The μC 11 accesses a non-volatile memory, here an EEPROM 26, via a function “CTL EEPROM” 24, from which the μC stores an operation parameter R
Load to AM25. The μC also stores the operating parameters in RAM 25 and EEPROM.
It can be stored in the OM 26. The μC 11 can transmit and receive data to and from the communication function COMM 28 via the bus interface 30. The received data can be used by the μC for motor control or stored in RAM 25 or EEPROM 26. The EEPROM 26 and the bus interface 30 are shown in FIG.

As a simple example, FIG. 1 shows an electronic commutation motor having only one phase winding 38. Such a motor is disclosed, for example, in DE 2346380C. The energization of the phase winding 38 is performed by a transistor stage (unit) 36. Outputs OUT1 and OUT2 of μC11 control npn transistors 141, 142, 143 and 144 connected as H bridge 37. OU
Current flows through the stator winding 38 in one or the other direction, depending on whether T1 is HIGH and OUT2 is set LOW or vice versa. Of course, the invention is likewise suitable for any type of electronic commutation motor, for example a three-phase motor. This is only an example.

The commutation is performed electronically. For this purpose, the position of the permanent magnet rotor 39 is detected via a Hall sensor 40, processed into a signal HALL via an electronic Hall circuit 41 (this will be described in detail in FIG. 3), and a drive function unit (unit) Further supplied to the AF 42. This drive function unit performs a hall interrupt routine HIR (
8 and 9), timer interrupt routine TIR (FIG. 11), ignition angle calculation routine ZWR
(FIG. 10), and a timer CNT_HL. The timer CNT_HL is a component of the μC11 used in this embodiment, but may be a separate component. The timer is used to measure time with high accuracy, and can be controlled by the instruction of μC11.

The drive function 42 provides accurate commutation of the transistor stage 36 and reliable driving, for example, when the transistor stage 36 is overloaded. The commutation when no ignition angle shift is performed is shown in FIG. The commutation for performing the ignition angle shift is shown in FIGS. 6 to 14B and FIG.

The rotation speed controller RGL 43 controls the motor rotation speed in the embodiment. (Of course, the motor M can be driven without the rotation speed controller 43.) The rotation speed can be controlled by, for example, a pulse width modulation generator (PWM generator) 34 or a block control (Bloc).
ksteuerung). Block control is shown at 60 by a dashed line. As for the block control, for example, DE 4441372.6 (intern
: D183i). This publication gives an example of such a block control.

The PWM generator 34 has a triangular wave generator 35, a control voltage generator 45, and a comparator 120, and is shown in detail in FIG. The present invention can of course be applied to an ECM that does not perform the rotation speed control.

The current limiter “I <Imax” 44 alleviates the conduction of the output stage 36 when the current becomes excessively large in only one phase winding 38. This is, for example, when starting the motor. The current limiter 44 is shown in detail in FIG.

At the end of the description, reference is made to advantageous values for the electronic components used in the individual figures of the examples.

FIG. 2 shows the COP842CJ from National Semiconductors used in the examples.
1 shows a pin arrangement of a microcontroller (μC) 11. The description in μC11 corresponds to the description of this manufacturer, and the description of each outer line is exclusively the reference used in this description. A quadrant is marked at the top left to indicate location and is shown in subsequent figures.

FIG. 3 is a detailed circuit diagram of components of the Hall circuit 41. This Hall circuit processes the signal of the Hall sensor 40. Furthermore, clock input terminals CK0 and C
The connection of K1 and the connection of the reset input terminal RES are shown. Other components are not shown in FIG.

A crystal oscillator 97 is connected to the terminals CK 0 and CK 1 of the μC 11 (see FIG. 3). This crystal oscillator sets a clock frequency of, for example, 10 MHz. The reset input terminal Res (FIG. 3) is connected to ground via a capacitor 99 and to + Vcc via a resistor 101. These two components form a power-up reset at switch-on as usual.

The hole generator 40 is connected to + Vcc and ground 100 via a resistor 106 for supplying current. The output signal uH is supplied to two input terminals of a comparator 110, and a filtering capacitor 110 is assigned to a Vcc input terminal of the comparator. An output terminal of the comparator 108 is connected to a non-inverting input terminal of the comparator 108 via a feedback resistor 112 and a so-called pull-up resistor 1.
14 and + Vcc. Further, the output terminal of the comparator 108 is directly connected to the port HALL (FIG. 3) of the μC 11. Thus, a rectangular signal HALL controlled by the rotor magnet 39 (FIG. 2) is obtained at this port.

Signals and Commutation of Hall Sensor 40 FIG. 4 is a diagram of the signal HALL (FIG. 3) and the associated commutation when “ignition angle shift” is not used. Commutation is directly controlled by the signal HALL, since no ignition angle shift is used.

The signal HALL ideally has the value HALL during a rotor rotation of 180 ° el.
= 0 and has the value HALL = 1 during the subsequent 180 ° rotation. HA
Each change from LL = 1 to HALL = 0, or vice versa, affects (acts on) the interrupt process in μC11. This interruption process is performed by the HALL-INT shown in FIG.
Is indicated by Y.

The time between two Hall changes, for example between t_O and t_E, is
It is called the hole length HL or the hole time t_H, and is shown as the actual hole length HL in FIG. The hole length is a measure for the number of rotations of the motor. The shorter this is, the higher the rotational speed of the rotor 39 (FIG. 1). (The actual value is the instantaneous value measured by the motor.)

In this embodiment, energization of the stator winding is controlled by output signals OUT1 and OUT2 of μC11 (FIGS. 1 and 2). These signals are shown in FIG. 4 as an example at a low rotation speed, and are also shown in FIG.

If OUT1 is 1 (HIGH) and OUT2 is 0 (LOW), the current is (FIG. 1) from the plus voltage _UN via the transistor 144, the stator winding 38, the transistor 141 and the measuring resistor 140. Flows to the ground.

On the other hand, if OUT1 is 0 and OUT2 is 1, the current (FIG. 1) _flows from the positive voltage UN through the transistor 142 through the stator winding in the opposite direction,
It flows to the ground via the transistor 143 and the measuring resistor 140. Stator winding 38
In this case, the current is conversely applied.

If the ignition angle shift is not performed, the point where the signal HALL changes, that is, H
At the ALL interrupt Y, the two values OUT1 and OUT2 are briefly zeroed by μC11. This is, for example, 50 μs, and all four transistors 141 to 141
44 is blocked for a short time, thereby avoiding a short circuit at the bridge 37. This is shown in FIG.

A simple hole interrupt for the commutation in FIG. 4 is shown later in FIG.

FIG. 4 shows the relationship between the rotation speed and the hole length . The relationship with the rotation speed n is shown below. This relationship is a function of the number of poles P of the rotor 39.

If the hole length HL ′ is measured in seconds, HL ′ = T / P (1) holds. Where T = duration of rotor rotation (seconds) P = number of poles of rotor 39.

If the number of rotations is measured at the number of rotations / min, HL ′ = 60 / (n × P) (2) Here, n = the number of revolutions per minute, and P = the number of poles of the rotor 39.

Since the hole length HL exists in μs in this embodiment and HL ′ is seconds, HL ′
To HL. HL = 1,000,000 HL '(3) P = 4, that is, for 4 poles, HL = 15,000,000,000 / n (4) Conversely, if P = 4, n = 15000000 / HL (5) where n = rotation per minute Number, HL = hole length at μs.

The number of revolutions of n = 2870 min ⁻¹ corresponds to a hole length HL of HL = 150000000/2870 = 5226 μs in the case of a four-pole rotor, for example.
In hexadecimal representation inside the processor, this is 0x146A (hexadecimal is indicated by the leading digit 0x).

Ignition Angle Shift In the motor of FIG. 1, the rotor position sensor 40 is arranged at the pole gap of the stator, ie, 0 ° el, and the change of the signal HALL is 0 ° el, 180 ° el.
, 360 ° el or the like. This is shown by way of example in FIG. Such a configuration of a Hall generator is described by way of example in DE-A 197 47 9.2 (In
tern: D201i), see FIG. 1, part (or member) 25.

However, when the motor rotates at high speed, optimization of output and efficiency is necessary.
Commutation of the current in the stator windings 38 should be performed before the change in the Hall signal. That is, in FIG. 4, it should be executed before t_0 and t_E in time. This can be referred to as early ignition or advanced ignition. For this purpose, the rotor position sensor 40 could be shifted relative to the stator of the motor 39. This is not useful, however, since the motor should normally rotate in both directions and pre-ignition should increase with increasing speed in both directions.

Therefore, the ignition angle shift is electronically controlled. For this purpose, the already described 16-bit timer CNT_HL (FIG. 1) is used. The timer CNT_HL is loaded with the start value t_TI (described above) at each hole interrupt Y, followed by the value 0
Counted down until When it reaches zero, the timer CNT_HL
At 11, a so-called timer interrupt is triggered, and the timer is automatically post-loaded (additionally loaded) with the content t_AR of the so-called auto-reload register AR (also in the μC 11) and starts anew. See S302 in FIG.

During hole interrupt Y, timer CNT_HL reaches zero at the point at which commutation is to be performed, thereby adjusting to trigger an interrupt. This timer interrupt is shown in FIG. 5 by T _N , T _{N + 1,} etc., and the hole interrupt is shown by H _N , H _{N + 1,} etc.

The operation of the timer is set by the μC used. This μC includes a timer. Here, in some cases, the timer can also be configured via a μC register. Possible configurations are, for example, triggering an interrupt when zero is reached, or automatic reloading of the timer when zero is reached.

In addition, a timer CNT_HL can advantageously be used here for measuring the hole length HL (FIG. 4). This is indicated by t_H _{N in} FIG.

FIG. 5 shows the calculation of the timer start value t_TI. Shown is the signal HA
LL, hall interruption H _N-1 , H _N, etc., timer interruption T _N-1 , T _N, etc., hall length t_H _N-1
, T_H _N, etc. The signal HALL is applied to the input Hall (FIG. 2) of the μC 11. In the embodiment, the hole lengths t_H _N−1 , t_H _N, etc.
/ 4 rotation, that is, the time required for 180 ° el.

The concepts of the hole length HL and the hole time t_H are used synonymously below. The hole time t_H _{N + 1} starts after the hole interruption H _N and does not include it, and ends with the subsequent hole interruption H _{N + 1 including} it. Hole and timer interrupts are numbered according to the hole time in which they occur. That is, the timer interrupt T _N and the hole interrupt H _N at the end of this time belong to the hole time t_H _N.

In FIG. 5, the value of the timer CNT_HL is plotted below the signal HALL.
The timer CNT_HL counts down between the respective values. For example, t_TI
From time t_0 to time t_E and time 312 from t_TI to t_E.

The timer start value t_TI for the hall time t_H _{N + 2} is calculated from the hall length t_H _N in this embodiment. This is shown symbolically in 300,
During the hall time t_H _{N + 1} , the value t_TI is calculated according to the following equation: t_TI: = t_H _N −t_ZW (6) That is, the (constant) ignition angle time t_ZW is subtracted from the hall length t_H _N.
Similarly, t_TI for the hole time t_H _{N + 3} is calculated from the hole length t_H _{N + 1} . This is symbolically shown at 301.

In this way, commutation is performed at time points T _N , T _{N + 1} , T _{N + 2,} etc. T _N is earlier than H _N by time t_ZW. That is, commutation is advanced. Similarly, T _{N + 1} is earlier than H _{N + 1} . The instants T _N , T _{N + 1,} etc. are indicated by upward pointing arrows.

Referring now to FIGS. 19 to 21, how very advantageously, for example in the four-pole rotor 39, the commutation time T _{N + 4} is determined by the temporally preceding hole length t_H _N A description will be given of whether a particularly quiet rotation of the motor can be obtained. This deformation is indicated by 304 in FIG.
Symbolized by 308. Similarly, in the case of a six-pole rotor, the commutation time T _{N + 6} is determined by the hole length t_H _N.

FIGS. 6 and 7 show two possible cases that can occur when measuring the hole length t_H with the timer CNT_HL.

Shown is the signal HA applied to the input Hall (FIG. 2) of the μC 11.
LL, hole interrupts H _N and H _{N + 1} , and timer interrupt T _{N + 1} (FIG. 7). The time axis of FIG. 7A shows the start value t_B and stop value t_E of the timer CNT_HL. These values are the hole length t, which is executed for the first time during the next hole time t_H _{N + 2.}
Used in the calculation of _H _{N + 1} . t_B corresponds to the start value t_TI timer CNT_HL during Hall interrupt H _N (the previously described), T_e corresponds to the stop value of the timer CNT_HL during Hall interrupt H _{N + 1.}

Two cases can occur.

In the first case (FIG. 6), the motor accelerates strongly and the hole interrupt H _{N + 1}
This is the case when it occurs before NT_HL reaches the value 0. In this case, the stop value of the timer CNT_HL is stored in t_E in the hall interrupt routine triggered by the hall interrupt H _{N + 1} (S202 in FIG. 8A), the motor is commutated, and the timer CNT_HL and the auto reload register AR are newly stored. The hall length t_H _N-1
Is loaded (FIG. 5), and the timer CNT_HL is newly started (S238 in FIG. 9B). Therefore, in FIG. 6, the timer interrupt T _{N + 1} does not occur during the hall time t_H _{N + 1} .

In this case, the hole length t_H _{N + 1} is calculated according to the following equation: t_H _{N + 1} : = t_B−t_E + t_CORR (7) where t_CORR is a correction value described in detail in S258 of FIG. This is shown in FIG. 6B.

In the second case (FIG. 7A), the timer CNT_HL reaches 0 before the occurrence of the hole interrupt H _{N + 1} . When it reaches zero, the timer interrupt T _{N + 1} shown in FIG. 11 is triggered. The value t_AR is automatically reloaded from the auto-reload register AR (FIG. 1) to the timer CNT_HL at the time of the timer interrupt T _{N + 1} and newly started. See S302 in FIG. t_B now has the same value as t_TI, and thus t_B
Corresponds to _AR.

This is shown in FIG. 7B. H_NFrom the point immediately after_{N + 1}Until the timer CN
T_HL counts down from t_B to 0, and when the value is 0, the timer interrupt T_{N + 1}Trigger
I do. At the start of this interrupt, a new t_B is loaded into the timer CNT_HL,
See FIG. 11, S302, and H_{N + 1}Count down anew until the time.
However, in this case, the value 0 is not reached, only the value t_E. Hall interrupt H _{N + 1} At this time, a new value t_B 'is loaded into the timer CNT_HL, and
The sheer is repeated.

In the timer interrupt routine called by the occurrence of the timer interrupt T _{N + 1} , commutation is performed as long as the ignition angle shift is turned on. FIG. 11, S31
8, see S320 and S322. And the flag KD (commutation execution) is 1
Is set to See FIG. 11, S324.

The following hole interrupt H_{N + 1}The timer CNT_HL is newly stopped.
And the end time t_E is stored. See FIG. 8A, S202. Hall length t_H _{N + 1} (FIG. 7) is based on the set flag KD (FIG. 10, S252).
In steps S254 and S258 of 0, it is calculated as follows: t_1: = t_B−t_E (8) t_H_{N + 1}= T_B + t_1 + t_CORR (9)

Here, t_1 is the time between the timer interruption T _{N + 1} and the hole interruption H _{N + 1 as shown} in FIG. To calculate the hole length t_H _{N + 1} , the value t_B must be added to the value t_1. This is because the timer CNT_HL counts down this value to zero between the hole interruption H _N and the timer interruption T _{N + 1} .
Further, in some cases, a correction value t_CORR of, for example, 40 μs is added. This is shown in FIG. 7B and will be described in more detail later in FIG. 10, S258. After the hole interruption H _{N + 1} and the calculation of the number of revolutions (S274 in FIG. 10), the flag KD must be reset again (KD: = 0, see S272 in FIG. 10).

In the numerical example H _N for FIG . 7 , the timer CNT_HL is set, for example, to the value t_TI = t_B = 9800 (previously calculated in step 303 of FIG. 5). Thus, t_B has the value 9800 μs during the calculation. At T _{N + 1} , when the timer CNT_HL reaches the value 0,
When a timer interrupt occurs, 9800 is newly loaded and started (S302 in FIG. 11). At H _{N + 1} , the timer CNT_HL reaches the value t_E = 9640.
The value t_CORR is 40 μs. Next, according to the equations (8) and (9), t_1: = 9800-9640 = 160 μs t_H _{N + 1} : = 9800 + 160 + 40 = 10000 μs Therefore, the hole length t_H _{N + 1} is 10,000 μs in this example, and n_i
= 1500000 / t_H _{N + 1} = 1500000/10000 = 1500 revolutions / minute (equation 5; 4 pole rotor).

Then, immediately after H _{N + 1} , the timer CNT_HL is loaded with a new value t_B ′. This value corresponds to the value (previously calculated) t_TI '. Step 30 in FIG.
See 0.

FIGS. 8A and 9B show a flowchart of an advantageous embodiment of an advantageous hole interrupt routine. That is, it is a flowchart of an interrupt routine depending on the rotor position. This interrupt routine is triggered when a predetermined rotor position is reached, and the hall length t
_H _N is detected. Further, if commutation is not executed in the timer interrupt routine, commutation is also executed. All registers and variables described below are 16 bits in size in the preferred embodiment.

In S302, the timer CNT_HL is stopped, and the stop time of the timer CNT_HL is stored as t_E.

In subsequent steps S 204 to S 208, the edge for the next hole interrupt is adjusted by μC 11. For this purpose, it is checked in step S204 whether HALL = 1. If it is 1, the edge to be triggered by the next hole interrupt is set to a falling edge (HIGH → LOW) in S206. Otherwise, in S208, the edge is set to the rising edge (LOW → HIGH).

In the next S210, two cases are distinguished based on the flag DE (rotational speed reached): If DE = 1, no timer interrupt occurs, or a timer interrupt occurs,
And the ignition angle shift is turned on. Both are signs that the motor has reached its speed, as will be explained in detail later. If DE = 0, the ignition angle shift has been interrupted (SZW = 0), and a timer interrupt has occurred. This is a sign that the minimum rotation speed n_min at which the ignition angle shift is turned on has not yet been reached, as will be described later.

For the case of DE = 0, commutation is performed and the timer CNT_HL
Is set to a fixed value t_max (maximum hole length). This fixed value corresponds to the minimum rotation speed n_min. For example, if the minimum number of rotations is 300 rotations / minute, equation (4)
Therefore, t_max = 1500000/300 = 50000 μs.

For this reason, OUT1 and OUT2 are set to 0 in S312.

In S214, the auto-reload register AR and the counter CNT_HL are set to t_ma.
x (for example, 50,000). The timer CNT_HL operates with a resolution of 1 μs in this embodiment. Setting CNT_HL to a length of 50,000 μs corresponds to a rotation speed of 300 rotations / minute. Based on this, the timer CNT_HL
Is started.

In S216, the flag DE (which was 0) is set to 1, and S218 to S222
Commutation is performed. If HALL = 1 in S218, S220
OUT1 is set to HIGH, otherwise, OUT2 is set to HIGH in S222. Switching off the ports OUT1 and OUT2 in S212 and S22
The program requires a certain amount of time for the program steps S214-S218 to be executed between the switching on of OUT1 and OUT2 at 0 and S222. Therefore, a sufficient commutation gap (FIG. 23: t_G) is maintained. This is, for example, 50 μs.

Then, the hole interrupt is left in S224.

If DE = 1 in S210, calculation of the hole length t_H and a new timer value t_TI are requested for the ignition angle shift in S230. The main program is formed by the function manager, which will be described in detail with reference to FIG. The function manager requests the routine by setting the flag and resets the request by resetting the flag. S23 to request calculation
When 0, the flag FCT_ZWV is set to 1.

In another possible embodiment for S230, the calculation is performed by a hole interrupt routine (FIG. 8).
Run directly with This is indicated by S232. If the calculation is performed in S232, the hall time t_H _N-1 (eg, t_H ₄ ) can be used for calculating the timer interrupt time t_TI belonging to the hall time t_H _N (eg, t_H ₅ ). If S230 is used, the hole time t_H _N-2 (eg, t_H ₃ ) is used, or an earlier hole time is used as shown in FIGS. If the calculation is performed in the hole interruption (S232), S230 is omitted. The following description relates to an embodiment without S232.

In S234 (FIG. 9B), it is checked whether the flag KD = 1 (KD = KD).
Commutation). If KD = 1, then a timer interrupt has occurred within the hall time belonging to the hole interrupt (as shown for H _{N + 1 in} FIG. 7A).
, And the ignition angle shift is turned on. In this case, the commutation has already been executed by the timer interrupt (T _{N + 1 in} FIG. 7A), and the process directly jumps to S238.

If KD = 0 in S 234, no timer interrupt has occurred within the hall time belonging to the hole interrupt. That is, the situation shown in FIG. Commutation gap (
23 is started by setting the two ports OUT1 and OUT2 to zero in S236. That is, no energy is supplied to the stator winding 38 (FIG. 1) for a short time. However, if a timer interrupt occurred but was not commutated in this interrupt because the ignition angle shift was inactive, the DE
This is taken into account in the branch below S210 for = 0 (FIG. 8A).

In S 238, the value t_TI calculated by the ignition angle calculation (FIG. 10 or FIG. 21) described later is loaded into the auto reload register AR and the timer CNT_HL, and the timer CNT_HL starts.

In S240, the ignition angle shift is set to be active by setting the flag SZW: = 1. This is because, in this case, the required rotation speed has been reached, for example, 300 rotations / minute.

In S 242, whether or not commutation has already been performed is checked based on the flag KD (commutation execution). If not done (KD =
0), it is checked based on the signal HALL in S244 whether OUT1 is set to HIGH in S246 or OUT2 is set to HIGH in S248. Here, the commutation gap (t_G in FIG. 23)
1 and OUT2 are turned off (S236), and steps S238 to S24
4 between the switch-on and the switch-on.

In step S250, the routine exits the hole interrupt routine.

FIG. 10 shows a flowchart of an exemplary routine for calculating the ignition angle.
This ignition angle calculation is performed in each hole interruption routine (FIG. 8) when the minimum rotation speed is reached.
Requested by setting the request bit FCT_ZWV (FIG. 17). S2 in FIG. 8A
See 30. The firing angle calculation is used when relatively high priority tasks are not required.
Called by the function manager 190 (FIG. 17). Therefore, it is not possible to say exactly when this calculation is performed. Therefore, the time point B _N at which the ignition angle calculation is performed
(E.g., FIGS. 13A and 14B) are not set accurately and show example time points.

It should be noted that the calculation of the hole length t_H always applies for the preceding hole time. For example, during the hall time t_H _N , the hole length t_H _N−1 is calculated.

In S 252, based on the flag KD, it is checked whether or not commutation has been executed by a timer interrupt (for example, T _{N + 1 in} FIG. 7). See S234 in FIG. 9B. If affirmative (KD = 1), according to S254, the hole length t_H is obtained from the start time t_B and the time t_1 as shown in FIG. 7 and described therein. This time t_
1 is the difference between t_B and t_E. If not (KD = 0), the hole length t_H is obtained from the difference between t_B and t_E according to S256. See FIG.

In S258, a correction time t_CORR is added to the hole length t_H. This means that the timer CNT_HL is reset at the start of the hole interrupt (FIG. 8A and FIG.
02 and then start for the first time in S232. Until then, a predetermined time added as t_CORR (for example, 40 μs) is necessary for the hole interrupt routine to obtain an accurate hole length t_H in S258.

At S260, the instantaneous hole length t_H is stored in the actual hole value t_i. The instantaneous actual Hall value is thereby used as a measure for the instantaneous rotational speed in all other program parts (eg control).

At S 262, the instantaneous start time of the timer CNT_HL is stored as t_B. This time can then be used in the next hole time for the calculation of t_TI.

An inspection of the number of revolutions is performed. This is because the ignition angle shift should be performed from a predetermined minimum number of revolutions, for example, 300 revolutions / minute. For this reason, in S264, t_
It is compared whether H> t_SZW. t_SZW (eg, 49664 μs, which corresponds to 0 × C200) is the maximum hole length before the ignition angle shift should be performed. If t_H is greater than t_SZW, the motor is excessively slow and the ignition angle shift is interrupted by SZW: = 0 at S266.

In S268, the commutation time t_TI, that is, the time at which the timer interrupt should be triggered is calculated. For this purpose, the value t_ZW, that is, the time to shift the commutation time earlier is subtracted in S268. This time is, for example, 200 μs
It is. This value may be constant, or may be a value depending on the motor parameter. This value t_ZW can be changed externally via the bus 30 (FIG. 16). t_ZW = 0
If so, the ignition angle shift is shut off.

Next, an ignition angle calculation routine is processed. Request bit FCT_ZWV (FIG. 17)
Is set to 0 in S270, and the flag KD is set to 0 again in S272.
As a result, the subsequent hall time can be used, and the motor control request bit FCT_RGL (FIG. 17) is set in S274, thereby requesting motor control.

Therefore, the main tasks of the ignition angle calculation routine of FIG. 10 are detection of the duration of the preceding hole length (S258), calculation of the commutation time for the subsequent hole time (S268), and control request (S274). It is.

FIG. 11 shows a flowchart for an example timer interrupt. This timer interrupt is used for motor control and is initialized by the preceding hole interrupt and is triggered when the started timer CNT_HL counts down to zero before the next hole interrupt is triggered. See FIGS. 7A and 7B.

When the value reaches 0, the timer CNT_HL is set to the auto-reload register A in S302.
The value t_AR of R is loaded and started anew. This is because the timer is also used for calculating the hole length t_HL at the same time. This step is automatically performed by the μC11 on this counter when it reaches 0 and has been included in this flowchart for clarity.

In S 304, whether or not the ignition angle shift is active is checked based on the flag SZW. If it is not active, the motor is running slower than the minimum speed. This becomes apparent when the ignition angle shift is not active at the time of the timer interrupt, the auto reload register AR and the timer CNT_HL are set to the maximum hole length corresponding to the minimum rotation speed n_min in S214 of the hole interrupt routine. This is because it is set to t_max. Nevertheless, if the timer interrupt (T _{N + 1 in} FIG. 7) occurs before the hole interrupt (H _{N + 1 in} FIG. 7), the minimum rotational speed n_
The value does not reach min, the flag DE (rotation speed reached) is set to 0, and the timer interrupt routine is left in S308.

If the ignition angle shift is active (SZW = 1), S304 to S310
Jump to Here, the two ports OUT1 and OUT2 are set to 0 at the start of the commutation gap.

Steps S 312 to S 316 include a commutation gap of sufficient length (
A program loop that causes t_G) in FIG. 23 is formed. For this reason, S31
At 2, the delay value t_DEL is assigned to the timer DEL_CNT. This is, for example,
It is. In S314, the counter DEL_CNT is decremented by 1, and in S316, the DE
It is checked whether L_CNT has already reached the value 0, ie whether the delay loop has been completely processed.

If the result is negative, the process jumps back to S314 again, and the loop is continued. If, for example, 10 μs are required to pass through the loop, the above values provide a delay of 50 μs, during which time both ports OUT1 and OUT2 have zero output signals. This affects the commutation gap t_G.

Subsequently, as usual, as described in S218 to S224 of FIG.
Commutation is performed. If the hall value HALL = 1 in S318, S3
OUT1 is set to HIGH at 20, otherwise OUT2 is set at S322.
Set to HIGH. Therefore, commutation is not
This is executed at the interrupt, that is, before the hole interrupt. This is shown in FIG. _{N + 1} Time T before_{N + 1}It is.

In step S324, the flag KD (commutation execution) is set to 1, whereby the hole interruption and the ignition angle calculation routine can be identified. Based on this, the hole interruption routine is left in S326.

FIG. 12 shows, as an example, a signal HALL and a hole interrupt H when the motor of the present invention is started. _N And timer interrupt T_NIndicates the time point. Hall time t_H_NIs a hole interrupt
H_N-1And H_NThe time between and it becomes shorter and shorter. Because the motor accelerates
It is. No timer interrupt occurs during each hall time. In this example, t_H₂and
In the following Hall time, the ignition angle calculation is performed, but in this example, based on the acceleration of the motor,
Is the timer interrupt T₁, T_TenAnd T₁₁Only occurs. Because the rotation speed is t_H₈Or
This is because the first time it becomes constant to some extent.

FIGS. 13 (A) and 14 (B) show the progress of FIG. 12 on an enlarged scale, with additional explanations.

FIGS. 13 (A) and 14 (B) show, by way of example, the lapse of time when the motor of the present invention starts. This progress reveals a link to the hole interrupt, the ignition angle calculation and the timer interrupt.

The following variables are used in FIGS. 13 (A) and 14 (B): DE: flag “revolution reached” KD: flag “commutation performed” SZW: flag “start ignition angle shift” t_AR: auto Value of reload register AR (FIG. 1) CNT_HL: Timer for timer interrupt and hole length calculation t_E: Stop time (end time) t_H: Hall length (hole time) t_B: Start time (start time) OUT1: Port of μC11 for energizing the motor OUT2: Port of μC11 for energizing the motor

The signal HALL at the input of μC 11 is plotted. Hall length t_
H is between the hole interrupts surrounding it, for example, t_H ₂ = 40 ms is between H ₁ and H ₂ and t_H ₃ = 35 ms is between H ₂ and H ₃ . The hole interrupts are indicated by H _N , the timer interrupts by T _N , and the execution of the ignition angle calculation by B _N. Here, N is an index of the belonging hole interrupt t_H _N.

There are some important variables under the signal HALL. These variables are μC11
Used for programs that run on The time display is shown in ms for space reasons, but is processed in μs time inside the program. At the start of the motor, a number of variables are initialized (column INIT). t_TI and t_B are 50ms
Is initialized by This corresponds to a rotational speed of 300 revolutions per minute, from which the ignition angle shift is switched on in this embodiment. DE and KD are set to zero. This is because the required number of revolutions has not been reached at first. And SZW is also 0
Is initialized by This is because the ignition angle shift is interrupted.

[0101] the first time 50ms is loaded in the first Hall interrupt _{H 0} to the auto-reload register AR and timer CNT_HL, timer CNT_HL is started. Since Hall length T_h ₁ is 60 ms, the timer interrupt _{T 1} occurs before hole interrupt _{H 1.}

Since the ignition angle shift is interrupted (SZW = 0), only one value DE is set to 0 in the timer interrupt routine (S306 in FIG. 11). This indicates to the hole interrupt that the motor has not yet reached the minimum speed n_min. Because it is the maximum hole length t_max larger holes length T_h ₁ corresponds to the minimum rotational speed n_min. The timer CNT_HL is automatically loaded with an auto-reload value t_AR of 50 ms and starts.

The hole interruption routine (FIG. 8) is called by the hole interruption H ₁ . 40ms
Stop time t_E is secured. This stop time is determined by the timer CNT_HL
Is 10 m between the timer interrupt T ₁ newly set at 50 ms and the hall interrupt H ₁
Caused by the passage of s. Since a DE = 0, commutation is performed at H _1, 50 ms is loaded into t_AR and CNT_HL, timer CN
T_HL is started. DE is set to one. No calculations are required.

During the hole length t_H ₂ , the motor first reaches an average minimum speed of 300 revolutions / minute for the first time. Accordingly Hall interrupt H ₂ before the timer CNT_HL counts down to zero is triggered. Therefore, the timer interrupt T ₂ does not occur.

In the hall interruption routine at the time of the hall change H ₂ , the timer CNT_HL is set to 10 m
The stop time t_E of s is secured. DE holds its value DE = 1 since the timer interrupt is not generated between the hole length T_h _2. Thus, the hall interrupt routine identifies that the number of revolutions exceeds 300 revolutions / minute. In the hole interrupt routine, an ignition angle calculation routine (FIG. 10) is required, and the ignition angle shift is activated by SZW: = 1. Since in the hole length T_h ₂ not yet commutation (KD = 0), the commutation is performed at H ₂ in Hall interrupt routine. Auto reload register AR and timer CNT_
HL is loaded with the value t_TI initialized to 50 at the start of the motor. This is because the ignition angle calculation has not been performed yet. And timer CNT
_HL starts anew.

The calculation of the ignition angle shift during the hole length t_H ₃ is performed for the first time. Since the timer interrupt has not occurred (KD = 0), the Hall length T_h ₂ which is calculated during the Hall length T_h ₃ results from t_B = 50 ms and T_e = 40 ms. From this, if the ignition angle shift time is t_ZW = 0.2 ms, a timer interruption time of 39.8 ms occurs. Timer start time of the Hall time t_H ₃ is secured to t_B.

The hole interruption routine up to the hole interruption H ₃ proceeds similarly to the hole interruption routine up to the hole interruption H ₂ . This is because the motor accelerates further and the hole interrupt occurs before the timer CNT_HL reaches the value 0. Therefore, no timer interrupt occurs during this hall time. This also occurs at hole interruptions H ₄ , H ₅ , H ₆ and H ₇ . The ignition angle calculation routines B ₄ , B ₅ , B ₆ , and B ₇ are also called at the respective hall times.

In the hall time t_H ₈ , the motor finally reaches the target rotation speed of 1500 revolutions / minute. This rotation speed corresponds to a hole length of 10 ms. In this embodiment, the timer interrupt time t_TI for holes time T_h _N is always because is calculated from the Hall length T_h _N-2 during the hole time T_h _N-1, there is "delay" of the two Hall time.
That first hole time timer CNT_HL is started with the correct timer interrupt time t_TI is T_h _10. Because the hall time t_H ₈ is 10m
This is because it was the first hall time of s, and the result from the hole length calculation for the hall time of t_H ₈ is used for the first time at t_H ₁₀ .

During the hall time t_H ₁₀ , the ignition angle calculation B ₁₀ is executed again. Start value t_TI for the auto-reload register AR and timer CNT_HL during the Hall interrupt routine to the H ₉ was 9.8ms.

[0110] Thus shall trigger the timer interrupt T ₁₀ at 9.8ms after the Hall interrupt H _9. The value t_AR (9.8 ms) is automatically loaded into the timer CNT_HL, and the timer CNT_HL is newly started. The ignition angle shift is turned on (SZW = 1). As a result, commutation is performed at the timer interrupt (T ₁₀ ). Flag KD is set to 1, thereby to ignition angle calculation and subsequent holes interrupt routine at H _10, to indicate that the commutation is performed.

[0111] In the Hall interrupt routine of the Hall interrupt H _10, the stop value of the timer CNT_HL is secured to t_E, the ignition angle calculation routine request, auto-reload register AR and timer CNT_HL is loaded, the timer CNT_HL start Is done. Since commutation has already been made at the time of the timer interrupt T ₁₀ timer interrupt routine, not more commutation.

The subsequent hall time t_H ₁₁ and the like elapse in the same manner as t_H ₁₀ unless the actual rotation speed or the target rotation speed of the motor changes.

Motor Control FIG. 15 shows circuit parts important for controlling and driving the motor. Elements that are the same or have the same effect as in the preceding figures are given the same reference numbers and will not normally be described again.

The terminal assignment of μC11 is shown in FIG. Output terminal OUT1 of μC11
And OUT2 are npn transistors 141, 141 connected as an H-bridge circuit 37.
142, 143 and 144 are controlled.

The output terminal RGL of the μC 11 is connected to the capacitor 124 via the resistor 123. When RGL is set HIGH, the capacitor 124 is charged and R
When GL goes low, the capacitor is discharged, and when RGL is TRISTATE, capacitor 124 is disconnected from RGL and maintains its voltage. Without the current limiter 44 described further below, the point 125 could be directly connected to the non-inverting input terminal (+) of the comparator 120.

If the npn transistor 150 is non-conductive, that is, if the current limiter 44 is inactive, the same voltage as the capacitor 124 is adjusted to the relatively small capacitor 127 via the resistor 126. Therefore, the voltage at the non-inverting input terminal (+) of the comparator 120 can be controlled via the output terminal RGL of the μC 11.

The triangular signal generated by the triangular wave oscillator 35 is applied to the inverting input terminal (−) of the comparator 120. The triangular wave oscillator 35 has a comparator 130. A positive feedback coupling resistor 132 is connected from the output terminal P3 of the comparator 130 to its non-inverting input terminal (+). Similarly, a negative feedback coupling resistor 131 is connected from the output terminal P3 of the comparator 130 to the inverting input terminal of the comparator 130. Have been.
A capacitor 135 is connected between the inverting input terminal of comparator 130 and ground 100. The output terminal of the comparator 130 is connected to the +
Vcc. The non-inverting input terminal of the comparator 130 has two resistors 1
It is connected to + Vcc or ground 100 via 34 and 136.

For the operation of the triangular wave generator 35 and the control of the output RGL of the μC11 by the μC11, refer to DE 198 682.8 (Intern: D216).

The voltage of the triangular signal at the inverting input terminal of the comparator 120 is
20 is lower than the voltage of the reference signal at the non-inverting input terminal of comparator 20,
The output terminal OFF of 0 is HIGH, and the lower transistors 141 to 143
Can be switched on and off by OUT1 and OUT2 via logical AND elements 147 and 148. If the voltage of the triangular signal is above the reference signal, the output OFF of the comparator 120 is LOW, so that the stator winding 38 cannot be energized.

The so-called duty ratio is adjusted via the voltage of the capacitor 124 and, consequently, the voltage of the capacitor 127. The duty ratio is the ratio of the duration during which the output of the comparator 120 is HIGH during the period of the triangular signal to the entire period. This duty ratio can be between 0% and 100%. If the motor speed is, for example, too high, the capacitor 124 is discharged via the RGL, thereby reducing the duty ratio. This is generally referred to as pulse width modulation (PWM). The pull-up resistor 128 turns off the open collector output terminal of the comparator 120.
Is used to raise the voltage to + Vcc at the time of HIGH.

In order to be able to start the motor upon switching on, the capacitor 124 is charged via the RGL for a predetermined duration during initialization. This allows the voltage at the capacitor 127 to reach the comparator 120 and thus the bridge 37.
Reaches the minimum required to switch on.

The current limiting section 44 is realized as follows. That is, the current is
And flows to the ground 100 via the measuring resistor 140. The higher the current through resistor 140, the higher the voltage at that resistor, and thus the point 1
The potential of 29 is also high.

When the potential of 149 reaches a predetermined value, transistor 150 is turned on and the voltage of capacitor 127 is reduced, whereby the duty ratio at the output terminal of comparator 120 is reduced. Resistor 126 also discharges large capacitor 124 during current limiting, which prevents it from promoting current limiting. This is because the small capacitor 127 can discharge more quickly. When the current limiting operation is completed, the relatively small capacitor 127 is replaced with the large capacitor 12.
4 and again at that voltage. Therefore, the resistor 126 and the capacitor 127 act so that the current limiter 44 has a higher priority than the control.

The current limiting section 44 has a filter element including a resistor 151 and a capacitor 152 for grounding. This filter element is operated by an npn transistor 150 which pulls the non-inverting input terminal (+) of comparator 120 to ground 100 when the voltage at its base is high enough. On the rear side, a filtering element including resistors 153 and 155 and a capacitor 154 is arranged.

For another form of the current limiter, see DE 19826458.5 (Intern: D215).
Please refer to. This current limiter can be formed by a comparator and can be program-controlled.

EEPROM Function FIG. 16 shows a part of a circuit relating to the EEPROM 26 and the bus interface 30. The pin arrangement of μC11 is shown in FIG. The same or correspondingly acting parts are provided with the same reference symbols as in the preceding figures. The EEPROM 26 is, for example, a type "2-Wire Serial CMOS EEPROM AT24C01A" (ATMEL).

The EEPROM 26 has a signal ESDA of μC11 (FIG.
) And receives the signal ESCL at its input SCL. The two lines are connected to + Vcc via resistors 172 and 173.

The write protection input terminal WP of the EEPROM 26 is connected to the pin CS (chip select) of the μC 11. If CS is HIGH, the EEPROM 26 is write protected, and if CS is LOW, data can be written to the EEPROM 26. The terminals VSS, A0, A1 and A2 of the EEPROM 26 are connected to the ground 1
00, and the terminal VCC of the EEPROM 26 is connected to + Vcc.

Accordingly, the lines ESDA and ESCL are serial buses between the μC 11 and the EEPROM 26. This bus is here driven as an IIC bus.

Normally, the EEPROM 26 is programmed once via the bus interface 30 at the factory. But reprogramming is always possible. Alternatively, the motor can be driven without bus 30. In that case, the EEPROM 26
Prior to using the motor, it is programmed by a known device.

[0131] The bus interface 30 operates with the IIC bus. The bus has a data line DATA with a terminal 160, which is connected via a resistor 162 to the terminal SDA of the μC11. From the terminal SDA, the resistor 165 is connected to + Vcc, and the capacitor 167 is connected to the ground 100. Further, the terminal SDA is connected to the emitter of the pnp transistor 168, the collector of which is connected to the ground 100,
The base is connected to a terminal N16 of μC11 via a resistor 169.

The bus interface 30 further includes a clock line CLOC having a terminal 161.
K. This clock line is connected to a terminal SCL of μC11 via a resistor 163. From the terminal SCL of the μC 11, the resistor 164 is connected to + Vcc, and the capacitor 166 is connected to the ground 100.

The circuit having the pnp transistor 168 is used to connect the output terminal N16 and the input terminal SDA of the μC 11 to the bidirectional line DATA of the IIC bus.

A further description of the EEPROM 26, the bus interface 30 and its programming can be found in DE 198 26 858.5 (intern: D215).

The value in the EEPROM 26 can be changed by the bus interface 30. Therefore, the minimum rotation speed n_min at which commutation by the ignition angle should be applied
Can be changed by setting the value t_SZW in the EEPROM, and thus the setting (configuration) of the motor can be changed. Similarly, for example, the ignition angle time t_ZW can be changed.

Function Manager FIG. 17 shows a flow chart according to a possible embodiment of the whole program executed on the μC 11. After the fan is switched on, an internal reset is triggered at μC11. In S600, the μC11 is initialized. For example, if the parameter is EEPR
The data is transmitted from the OM 26 to the RAM of the μC 11.

After initialization, the process jumps to the function manager 190 described above. The function manager starts in S602. This function manager controls the individual subprograms and sets their priorities.

First, functions that are time-critical and must be processed when executing each loop are processed. The commutation function COMM in S602 belongs to this. This is because the IIC bus (FIG. 16) has 25
This is because the inspection must be performed every 0 μs.

FIG. 18 shows an example function register 195, which has one bit reserved for another function.

In this embodiment, the function register 195 is one byte large and defines the following request bits for the requestable functions described below, starting from the least significant bit (LSB): Bit 1: FCT_ZWV for ignition angle calculation routine Bit 2: FCT_RGL for any type of control routine

The remaining bits are reserved for additional requestable functions that can be added to function manager 190 as needed.

When a given requestable function is requested by another function or interrupt routine,
The bit of the function to be requested is set to one. Next time, if the function manager 190 does not call another requestable function having a relatively higher priority during the loop execution,
This function is performed.

If the requested function has been processed, this sets its bit (FIG. 18) to 0 again. For example, FCT_RGL: = 0.

In FIG. 17, after S602, starting from the most important requestable functions in a predetermined order, it is checked whether those request bits are set, respectively. If this is a function, the function is executed, and based on this, the process jumps to the start S602 of the function manager 190 again. The inspection order of the function register 195 sets the priority of the function that can be requested. The higher such a function is in the function manager, the higher its priority.

The processing time of the function to be called is determined by the function that is always executed (S60 in this case).
When added to 2) and the interrupt routine, the maximum allowed time between two queries on the IIC bus 40 must not be exceeded. In the above example where the baud rate is 2k and the maximum allowable time is 250 μs, the maximum processing time for the functions called in S610 and S614 is about 100 μs.

In S610, it is checked whether the request bit FCT_ZWV for the ignition angle shift is set, that is, whether it has the value 1. If it is set, the routine jumps after S612, and the ignition angle calculation routine (FIG. 10 or 21) is executed. Before the end, the ignition angle calculation routine resets the request bit FCT_ZWV, and requests the control routine by setting the request bit FCT_RGL in S274.

If FCT_ZWV is not set in S610, FCT_R is set in S614.
It is checked whether GL is set. If set, a control routine for controlling the motor speed is called in S618.

If the bit to be checked is not set in both S610 and S614, the process jumps to S602 again, and the function executed by the function manager 190 at the time of executing each loop is newly called.

FIG. 17 symbolically shows a hole interrupt at 620. This hole interrupt has the highest priority L
1 (level 1). The hole interrupt has this high priority because accurate detection of the hall signal is very important for the quiet running of the motor 39. The hole interrupt interrupts all processes of the function manager 190, as symbolically indicated by arrow 621.

A timer interrupt is shown at 622 below the hole interrupt. This timer interrupt has a lower priority L2 and suspends all processes below it, as indicated by arrow 623. Accurate commutation is likewise very important for quiet running of the motor. Therefore, timer interrupt 622 has the second highest priority.

If a hole interrupt and a timer interrupt are requested at the same time, they are processed according to their priority order.

The function COMM has the next lower priority L3. This is because data must not be lost via the bus 30 during commutation.

The function ZWV has the next lower priority L4. This function can be requested in S230 and is shown in FIG. 10 (or FIG. 21).

The function RGL (S614) has the lowest priority. This is because the speed of the motor usually changes slowly due to its mechanical inertia, so that the control functions are usually not critical in time. However, steps S610 and S614
Order, and therefore their priorities, can be swapped.

In this way, the various “needs” of the motor 39 are assigned to a predetermined hierarchy,
The resources of μC11 are ideally used for driving the motor.

FIG. 19 shows a quadrupole outer rotor 39 in consideration of the ignition angle shift in consideration of the magnetization error of the rotor 39. This rotor has radially magnetized poles 534, 535, 536, 537 which are separated from one another by transitional (boundary) regions 530-533 as shown. I have. As an example, a so-called trapezoidal magnetization is assumed. See FIG. 20A.

Due to the inhomogeneity of the magnetized material and unavoidable errors in the magnetizing device (not shown), the course of the magnetic flux density is not precisely defined, especially in the transition regions 530-533, and varies somewhat from rotor to rotor.

Assuming that the rotor 39 rotates in the direction of arrow 540 past the hall generator 40, the hall generator 40 provides a hall voltage uH. The course of this Hall voltage is exaggerated in FIG. 20A for clarity. The portion 534 'of the Hall voltage uH is formed by the rotor pole 534 (N pole) and is slightly shorter. In other words, the zero passage of the Hall voltage is 0 ゜ el and almost 170 ゜ el
It is. However, the desired angle is exactly 180 ° el, which is 0 ° el.

The portion 535 ′ of the Hall voltage is generated by the rotor pole 535. This part starts at 170 ° el and ends at approximately 370 ° el. That is, it is too long.

The portion 536 'is created by the rotor pole 536 and extends from approximately 370 ° el to approximately 550 ° el. That is, it has the correct length, but the phase position is not accurate.

The portion 537 'is generated by the rotor pole 537 and is approximately 550
It has grown to 20 ゜ el. That is, it is too short. 720 ° el again corresponds to 0 ° el in this motor. This is because the rotor 39 makes one complete rotation.
The voltage course is then repeated as shown at 534'A in FIG. 20A.

FIG. 20B shows the associated signal HALL. This signal reflects the magnetization error just described. That is, the first portion 534 ″ is excessively short and the second portion 53
5 "is too long, the third portion 536" is out of phase, and the fourth portion 537 "is too short. After an angle of 720 ° el, the portion 534" A begins, but this portion has a higher rotational speed. In certain cases, it corresponds to part 534 ".

The parts 534 ″ and 537 ″ are mistaken for an excessively high speed, and the parts 535 ″
"Is mistaken for an excessively low rotational speed.

If the portion 534 "is used to calculate the time t_TI for the portion 536", the portion 536 "is commutated too early as described in the previous embodiment.

If the part 535 ″ is used in the calculation of the time t_TI for the part 537 ″, it is commutated too slowly there.

This leads to irregular rotation of the motor, and the motor noise increases.

Thus, according to the invention, the hole length of the portion of the signal HALL is advantageously used for calculating the time t_TI for the portion after one rotor rotation. This is symbolically illustrated in FIG. 5 by reference numerals 304, 306, 308 for a four pole rotor. For example, in FIG. 20B, the hole length t_H _N of portion 534 ″ is used to calculate the time for portion 534 ″ A. This is a symbolic example of 542,544
, 546. Then such an error does not occur. This is because, for example, when the rotational speed is constant, the portion 534 "and the portion 534" A are equal, and therefore no error is added.

FIG. 21 shows the ignition angle calculation routine modified accordingly for the commutation due to the ignition angle shift. Here, a compensation for the magnetization errors of the rotor 39 takes place as described above. All components already appearing in FIG. 10 have the same reference numerals and will therefore not be described again. The reader should refer to the description there.

In step 268 ′, direct calculation of the timer start value t_TI (S26 in FIG. 10)
8), two variables t_4 and t_3 are additionally used to temporarily store the calculated timer start value t_TI. The timer start value t_4 calculated from the hole length t_H _N-4 is assigned to the timer start value t_TI used for the subsequent hall time t_H _N.

Subsequently, the calculated timer start value is shifted so that this timer start value is the correct variable for the next ignition angle calculation. The value t_3 calculated from the hole length t_H _N-3 is shifted after t_4, and the timer start value (t_H-t_ZW) calculated in the instantaneous ignition angle calculation is stored in t_3 (where t_H is the hole length t_H). _N-2 ).

Further, step S267 is newly inserted. The memory variables t_4 and t_3 are set to the value 50000 upon interruption of the ignition angle shift (S266: SZW: = 0), so that these variables have a unique state.

FIG. 22 shows an example of a hole interrupt routine for the commutation of the present invention when the ignition angle shift is not performed. This is similar to that shown in FIG. At each hole interrupt (Y in FIG. 4), the program just running is interrupted, the so-called environment (eg stack pointer and registers) of μC11 is stored, and the interrupt routine belonging to this interrupt is called. . When the interrupt routine is processed, this routine outputs an instruction RETI (Return From Interrupt). Based on this, the environment of the μC 11 is formed again as before, and the interrupted program is processed further.

In this embodiment, a 16-bit timer CNT_HL is similarly used for measuring the hole length HL (FIG. 4). The timer counts down continuously starting from a predetermined start value, and jumps to its maximum value again when it reaches 0 when it reaches 0. In this way, it operates like a ring counter. This timer is also a component of μC11. The hole length HL can be used here for speed control.

In step S702, the actual hole length HL (see FIG. 4) is detected. Actual timer value t_E
(FIG. 4) is read from the timer CNT_HL and stored the “old” timer value t
The hole length HL is calculated by subtracting _O (FIG. 4: at the time of the preceding timer interrupt Y). To this end, t_Et_O is calculated and the result is double complement (Zweierkomple).
ment) is formed. In this way, a correct counting difference is always obtained if the timer does not further elapse more than half of its maximum value.

Based on this, the instantaneous timer value t_E is stored in t_O (S702). The resolution of the timer CNT_HL used in this embodiment is 1 μs, so the hole length HL exists in μs units.

For example, if t_O = 45000 and t_E = 35000, the hole length HL = (
45,000-35000) = 10000, which corresponds to 10,000 μs.

In the next step, commutation is performed. In S704, HALL = 1 (
HIGH) is checked. If HALL = 1, OUT2 is L at S710.
Set to OW. Since OUT1 and OUT2 are now LOW, a temporal commutation gap is inserted in S712. This is to avoid a short circuit in the bridge circuit 37 during commutation. The commutation gap has a duration of, for example, 50 μs. At S714, OUT1 is set to HIGH. In step S716, the port HALL of the μC 11 is configured on which edge this can trigger the hole interrupt HALL_INT. This edge can be adjusted to trigger an interrupt on a HIGH to LOW transition (falling edge) or on a LOW to HIGH transition (rising edge). Since the Hall signal is HIGH in branches S710 to S716, the port H
ALL must adjust to the falling edge, ie, the interrupt on the transition from HIGH to LOW. This triggers a hole interrupt again at the next hole change. This is performed in S716.

If HALL = 0 (LOW) in S704, S720, S722, S724
In the same way, the opposite commutation is performed, and HALL_INT is set to the opposite in S726. In step S730, the hall interrupt routine is left according to FIG.

FIG. 23 schematically shows the commutation process for n> 300 revolutions / minute, for example, for 2000 revolutions / minute. Therefore, in this case, an ignition angle shift is performed.

FIG. 23A shows the rotor position signal HALL. This signal triggers a rotor position dependent interrupt (FIG. 8) at positions H _N , H _{N + 1} , H _{N + 2} respectively. That is, the hole interrupt is output at Y in FIG.

Starting from the hole interruption H _{N, the} time t_TI is measured by the timer CNT_HL. This time is calculated from the values t_HN and t_ZW according to equation (6). The value t_ZW can be changed via the bus 30 as already described.

At time T _{N + 1} , the timer CNT_HL reaches the value 0, and triggers the motor control interrupt routine, that is, the timer interrupt according to FIG.

According to S310 of FIG. 11, at time T _{N + 1} , both OUT2 (FIG. 21B) and OUT1 (FIG. 21C) are set to zero. That is, the winding 38 is separated from the current supply, and after the commutation gap t_G (program steps S312 and S314).
, S316) the signal OUT1 is set HIGH at S322. This is because HALL = 1. On the other hand, OUT2 remains LOW as stored in step S310. OUT1 = HIGH means that the transistors 141 and 144 of FIG. 1 are conducting.

Similarly, at time T _{N + 2} , two signals OUT1 and OUT2 are set to LOW in step S310 of the routine in FIG. Then, after the commutation gap t_G, the value OUT2 is set to HIGH. Because HA
This is because LL = 0. See steps S318 and S322 in FIG. In contrast, OUT1 remains at the value LOW stored in step S310. Thus, the transistors 142 and 143 of FIG. 1 conduct.

FIG. 24 shows the signal HALL at the bottom and the current i_ of the single stator winding 38 at the top.
M. In FIG. 24, the ignition angle shift is cut off. That is, t_ZW = 0. After the commutation, it can be seen that the current i_M changes slowly at the time point H _N (change of the signal HALL). Therefore, this current only reaches a small amplitude in this case,
That is, the output generated by the motor M is low.

FIG. 25 also shows the signal HALL at the bottom and the current i_M (FIG. 1) at the top. However, in this case, early commutation (early ignition) is performed. That is, the current i_M
Is commutated ahead of the hole change H _N by the time t_ZW. Current i
It is clear that _M changes very quickly immediately after commutation, reaching a much larger amplitude than in FIG. In this case, the motor M produces a relatively large output and can therefore reach a relatively high rotational speed. In the case of FIG. 25, the commutation is advanced by about 15 ゜ el from the change of the signal HALL.

A typical example for the values of the components used is shown in the table: Capacitor: 135 1.5 nF 127, 152 10 nF 99, 110, 166, 167 33 nF 154 100 nF Tantalum capacitor 3.3 μF Resistance: 140 3Ω 162 , 163 47Ω 153,155 1kΩ 133,136 2.2kΩ 106 3.3kΩ 164,165 4.7kΩ 123,131,132 10kΩ 172,173 22kΩ 114,126 33kΩ 134 47kΩ 101,112,128,169 100kΩ npn transistor BC846 pnp transistor 168 BC856B Comparator 108, 120, 130 LM2901D Hall sensor 40 HW101A EEPROM 26 2-Wire Serial CMOS EEPROM AT24C01 (ATE EL) micro-controller 11 COP842CJ (Nat. Semicond.)

[Brief description of the drawings]

FIG. 1 is an overview diagram as an example of an embodiment of the present invention.

FIG. 2 is a pin layout diagram of μC COP842CJ.

FIG. 3 is a circuit diagram showing components for processing a hall signal.

FIG. 4 is a diagram of Hall signals and commutation when ignition angle shift is not performed.

FIG. 5 is a schematic diagram for explaining that the (early) commutation time T _N is calculated based on a value derived from a signal HALL.

FIGS. 6A and 6B are diagrams showing hole length calculations when a timer interrupt is not performed.

FIGS. 7A and 7B are diagrams showing hole length calculations when a timer interrupt is present.

FIG. 8A is a flowchart of a hole interruption routine based on an ignition angle shift, together with FIG. 9B.

FIG. 9B is a flowchart of the hole interrupt routine based on the ignition angle shift, following FIG. 8A.

FIG. 10 is a flowchart of an ignition angle calculation routine.

FIG. 11 is a flowchart of a timer interrupt routine based on ignition angle calculation.

FIG. 12 is a diagram of a hall signal at the time of starting the motor.

FIG. 13 (A) is a diagram showing, together with FIG. 14 (B), a Hall signal and a variable to which a drive function belongs.

FIG. 14B is a diagram, following FIG. 13A, showing a Hall signal and a variable to which a drive function belongs.

FIG. 15 is a circuit diagram showing an important part for controlling and driving the electronic commutation type motor.

FIG. 16 is a circuit diagram showing an important part for controlling the EEPROM and for data connection via the bus 30.

FIG. 17 is a schematic diagram of an advantageous embodiment of the function manager.

FIG. 18 is a schematic diagram of a function register used in the function manager.

FIG. 19 is a schematic view of a permanent magnet of a four-pole outer rotor.

FIGS. 20A and 20B are schematic diagrams for explaining the effect of a magnetization error of the outer rotor of FIG. 19;

FIG. 21 is a flowchart of the ignition angle calculation similar to FIG. 10, but modified in an advantageous manner.

FIG. 22 is a flowchart of a hole interrupt routine for the commutation as shown in FIG. 4;

FIG. 23 is a schematic diagram of the commutation process for the case where the commutation point is electronically shifted earlier.

FIG. 24 is a diagram showing the elapse of time between the signal HALL and the current i_M of the motor winding. Here, the shift in the commutation time is not performed in the early direction.

FIG. 25 is a diagram showing the elapse of time between the signal HALL and the current i_M of the motor winding. Here, a shift in the commutation time is performed in an early direction.

[Other]

8A and 8B are changed to FIGS. 8A and 9B, and FIGS. 12A and 12B are changed to FIG. 13A.
) And FIG. 14B, and the other figures are correspondingly lowered. Corresponding descriptions in the description and the claims have also been corrected. The control table is shown below.

──────────────────────────────────────────────────続 き Continuation of front page (81) Designated country EP (AT, BE, CH, CY, DE, DK, ES, FI, FR, GB, GR, IE, IT, LU, MC, NL, PT, SE ), OA (BF, BJ, CF, CG, CI, CM, GA, GN, GW, ML, MR, NE, SN, TD, TG), AP (GH, GM, KE, LS, MW, SD, SL, SZ, UG, ZW), EA (AM, AZ, BY, KG, KZ, MD, RU, TJ, TM), AL, AM, AT, AU, AZ, BA, BB, BG, BR, BY , CA, CH, CN, CU, CZ, DE, DK, EE, ES, FI, GB, GD, GE, GH, GM, HR, HU, ID, IL, IN, IS, JP , KE, KG, KP, KR, KZ, LC, LK, LR, LS, LT, LU, LV, MD, MG, MK, MN, MW, MX, NO, NZ, PL, PT, RO, RU, SD, SE, SG, SI, SK, SL, TJ, TM, TR, TT, UA, UG, US, UZ, VN, YU, ZW (72) Inventor Hornberger, Jörg Germany D-72280 Dorundshutten Nottental 1 (72) Inventor Yeske, Frank Germany D-78112 Sankt Georgen Friedrich-Ebert-Strasse 13 (72) Inventor Rappenecker, Harman Germany D-78147 Ferenbach-Crankenhaus Strasse 26 ( 72) Inventor Calvert, Arno Germany D-78628 Rottweil Groshofenstraße 1 0 F term (reference) 5H560 BB02 BB12 DA02 DB14 DC01 DC12 EC02 GG04 JJ02 RR06 TT02 TT13 TT15 UA02 XA02 XA12 XA15

Claims (29)

[Claims] 1. An electronic commutation having a stator, a rotor (39) and a program-controlled microprocessor or microcontroller (11) used to control commutation of the motor. An apparatus for detecting an amount of time (t_H) substantially inversely proportional to the number of revolutions of a rotor (39), and for calculating a time (t_TI) dependent on the amount of time (t_H). The device and the motor control interrupt routine (FIG. 11) are executed at a time interval (t_
A device for triggering at TI), the time interval corresponding to a time (t_TI) dependent on the detected amount of time (t_H), the motor control interrupt routine Have an effect (
Bewirken). An electronic commutation type motor including a program step (S310, S318, S320, S322). 2. The motor control interrupt routine (FIG. 11) determines that the time (t_TI) that depends on the amount of time detected is the time (t_H) required for the motor to actually pass through a predetermined angular path. 2. The motor according to claim 1, comprising a program step (S304, S306) for preventing the commutation from acting if it is larger. 3. The motor according to claim 2, further comprising a device for triggering a rotor position-dependent interrupt routine (FIGS. 8, 9) at a predetermined rotor position. 4. A timer controllable by an interrupt routine (FIGS. 8 and 9) dependent on the rotor position to detect an amount of time substantially inversely proportional to the number of revolutions of the rotor.
The motor according to claim 3, further comprising: (CNT_HL). 5. A timer (CNT_HL) controls a motor control interrupt routine (FIG. 11).
5. The motor according to claim 4, wherein the motor is also configured to trigger: 6. A timer (CNT_HL) includes an interrupt (
8 and 9), a first predetermined count value can be loaded, said count value corresponding to a time interval (t_TI) that depends on the detected amount of time (t_H); 6. The motor according to claim 5, wherein a motor control interrupt (FIG. 11) acts after counting a predetermined count value. 7. An interrupt depending on the rotor position (FIGS. 8 and 9) is a motor control interrupt (FIG.
Motor according to one or more of claims 1 to 6, having a higher priority than in Fig. 11). 8. The timer (CNT_HL) has a motor control interrupt (FIG. 11: S
During the period 302), a predetermined count value (t_AR) can be loaded, and following the loading process, counting is performed until the next interrupt (FIGS. 8 and 9) depending on the rotor position, whereby the predetermined count By forming the difference between the numerical value (t_AR) and the count state (t_E) when the next interrupt (FIGS. 8 and 9) depending on the rotor position is reached, the time interval between the interrupting processes (FIG. 8). The motor according to any one of claims 4 to 7, wherein 7A: t_1) is detected. 9. An auto-reload register (AR) for loading the predetermined count value (t_AR), the auto-reload register storing a first predetermined count value (t_TI), and a timer 9. The motor according to claim 8, wherein (CNT_HL) is supplied as a predetermined count value during a motor control interrupt (FIG. 11). 10. A method for commutating an electronic commutation type motor depending on the number of revolutions, the motor having a stator and a rotor, further comprising a program control for controlling commutation of the motor. A) detecting the amount of time (t_H) substantially inversely proportional to the number of revolutions of the rotor; b) detecting the time; Calculating a count value (t_TI) from the quantity (t_H) according to a predetermined calculation rule; c) starting from a predetermined first rotor position, a first value corresponding to said calculated count value.
D) after the first time has elapsed, triggering commutation (T _N ); e) a subsequent second time until reaching a predetermined second rotor position ( t_1)
F) adding the first time and the second time and correcting or not correcting the sum by at least one correction factor, the amount of time substantially inversely proportional to the rotational speed of the motor (
t_H). 11. The predetermined calculation rule includes a subtraction step, wherein the amount of time (t_H) is substantially inversely proportional to the number of revolutions of the rotor.
11. The method according to claim 10, wherein a predetermined time (t_ZW) is subtracted from. 12. When the first time corresponding to the calculated count value (t_TI) is larger than the time interval between a predetermined first rotor position and a predetermined second rotor position, the two times are determined. The method according to claim 10 or 11, wherein the time interval between the rotor positions is directly detected and used as an amount of time (t_H) substantially inversely proportional to the rotational speed of the motor (S256). 13. A time value (t_H) substantially inversely proportional to the rotation speed of the motor is compared with a predetermined value (t_SZW) corresponding to the minimum rotation speed (FIG. 10: S264), and a logical value corresponding to the comparison result (SZW) is intermediately stored (FIG. 10: S266), and when the logical value (SZW) is a predetermined value, the trigger of commutation performed after the lapse of the first time (t_TI) is suppressed (FIG. 11). : S304, S30
6) The method according to any one or more of claims 10 to 12. 14. When a predetermined rotor position is reached, an interrupt depending on the rotor position is executed by an interrupt routine (FIGS. 8A and 9B), and a timer (CNT_HL) used for time measurement at the start of the interrupt routine. Is stopped (S202), and its instantaneous value is stored in a variable (t_E).
4. The method according to any one or more of three. 15. In an interrupt routine depending on the rotor position, a timer (CNT_HL) used for time measurement is provided with a count value (t_TI) calculated beforehand in accordance with a predetermined calculation rule in time after the stop (S202). 15.) The method according to claim 14, wherein the method comprises loading and starting (Fig. 9B: S238). 16. A method for detecting a time amount (t_H) substantially inversely proportional to the rotation speed of a motor by using a time interval between a stop and a start of a timer (CNT_HL) used for time measurement as a correction coefficient (t_CORR). The method according to claim 14 or 15, which is used for: 17. A count value calculated from a predetermined first rotor position and calculated (
t_TI) is determined by the amount of time substantially inversely proportional to the number of revolutions of the rotor, which is detected approximately one revolution before the actual rotation of the first time. 17. A method according to any one or more of claims 10 to 16, wherein the method is calculated from: 18. The method according to claim 10, wherein at least one method step which is not critical in time is configured as a subroutine (FIG. 10), said subroutine being called during the course of the program when processor time is available (FIG. 17). 18. The method according to any one or more of claims 17 to 17. 19. The calculation of the amount of time (t_H) substantially inversely proportional to the number of revolutions of the motor and the calculation of the count value (t_TI) which is the basis of the measurement of the first time are performed by the above-described subroutine. The method according to claim 18, (FIG. 10). 20. At least one parameter (t_ZW) required for calculation is obtained from a non-volatile memory (26) assigned to a motor and a microprocessor (11).
20. A method according to any one of claims 10 to 19, wherein the method loads into a RAM (25). 21. The nonvolatile memory (26) is associated with a bus (30), via which at least one parameter can be changed in the nonvolatile memory (26). Method. 22. An electronic communi- cation having a stator and a rotor (39) and a program-controlled microprocessor or microcontroller (11) used for controlling commutation of the motor (M). In the attraction motor, at at least one predetermined rotor position, a timer (CNT_HL) is started by a predetermined start value (t_TI), and the timer (CNT_HL) is activated after a lapse of time depending on the start value (t_TI). An interrupt (FIG. 11) is triggered in the program of the microprocessor (11), and commutation of the motor (M) during the interrupt (FIG. 11: S318, S320).
, S322) is performed. 23. A start value (t_TI) of a timer (CNT_HL) is a function of a time (t_H) that depends on a rotation speed, and the time is determined by rotating a rotor by a predetermined rotation angle in a time domain preceding commutation. 23. The motor of claim 22, which is the time taken to perform. A predetermined time (t) for calculating a start value (t_TI).
24. The motor according to claim 23, wherein _ZW) is subtracted from a time (t_H) dependent on the rotation speed. 25. A method for detecting a speed-dependent quantity in an electronic commutation motor (M), the motor comprising a stator, a permanent magnet rotor (39), and a current controlled by the rotor. A method which is of the type comprising a magnetic (galvanomagnetisch) sensor (40), a microprocessor or microcontroller (hereinafter microprocessor), a control program assigned to the microprocessor, and a timer (CNT_HL). A) converting the output signal of the galvanomagnetic sensor (40) into a rectangular signal (HALL); b) detecting a predetermined signal change of the rectangular signal (HALL) by a microprocessor, and controlling the rotor position by a control program. C) depending on the rotor position (Y: FIG. 4: Y) ), The first counting state of the timer (FIG. 4:
d) holding the second counting state of the timer in the event of a subsequent rotor position-dependent interrupt (Y); e) depending on the rotational speed from the difference between the two counting states. Detecting a value (HL in FIG. 4: HL) corresponding to the time required for the rotor (39) to rotate through a predetermined rotation angle (FIGS. 4 and 22). 26. A program-controlled macro-processor or micro-controller (11) used to control the commutation of the stator (38), the rotor (39) and the motor (M) (hereinafter microprocessor). ) And a rotor position generator (40, 41), the output signal of said rotor position generator being divided by said microprocessor for evaluation by said microprocessor (11). The input signal is supplied to an input terminal capable of inputting and processed by the microprocessor. At least one output terminal of the microprocessor outputs the rotor position generator (4
0, 41), the control signals (OUT1, O2) shifted by the shift time.
UT2) for commutation of the motor, wherein the duration of the shift time is a predetermined function of the number of revolutions. 27. A microcontroller (11) comprising a timer (
27. The electronic commutated motor (M) of claim 26, comprising at least one CNT_HL), wherein the timer controls at least one output of a microprocessor that functions as an output of a control signal. 28. The electronic commutation motor (M) according to claim 27, wherein the timer (CNT_HL) is automatically post-loaded with the value (t_AR) in a predetermined state, and the timer is newly started. ). 29. A microprocessor comprising: a rotor position generator (40, 41);
28. The electronic communi- cation of claim 26 or 27, wherein an interrupt is triggered on each change of the signal (HALL) of the motor, the timer (CNT_HL) and the interrupt being used for measuring the hole length (HL) in the motor. Stationary motor (
M).